Summary Multimodal sensory information is converged through poly-synaptic pathways to the hippocampus to form integrated representations and encode memories of the world, which in turn guide our future behavior through multiple poly-synaptic downstream pathways. It is well known that each of the brain regions in this circuit consists of various neuronal cell types or groups which are distinct in connectivity and functions. Yet we cannot construct a long-range poly-synaptic wring diagram among these specific neuronal groups without efficient tools to continuously track the multiple orders of synaptic connections in a controlled and directed manner. Similarly, it is not clear if the wiring of this pathway is subject to dynamic modifications by learning, memory or brain disorders due to a lack of tools to continuously monitor the neuronal connectivity over time. Here we propose to develop novel methodology to embrace these challenges. In the first part we will modify viral vectors for trans-neuronal tracing which can spread across synapses in a controlled, stepwise manner. With the new tool we will test the hypothesis that distinct neuronal groups in the CA1 or CA3 regions of the hippocampus receive distinct poly-synaptic inputs and send out distinct poly-synaptic outputs. In the second part we will conduct functional imaging and functional manipulations of the specific CA1 or CA3 neuronal groups to test the hypothesis that these neuronal groups, defined by their specific poly-synaptic inputs and/or outputs, receive distinct sensory information and in turn adjust different aspects of behavior. In the third part we will develop novel technology to trace in the same animal the connectivity of neurons of interest at two time points--before and after a learning process--to examine if learning and memory alters neuronal connectivity. Together, these studies will demonstrate whether the distinct neuronal groups at each brain region in the memory circuit are selectively connected with specific neuronal groups in other brain regions to form functional ?channels? which bridge different sensory information to the different aspects of behavior. The novel tools to be developed/optimized in this study will find wide use in neuroscience research helping us to map out the functional network in the brain.